When do extraction rights help forests? Robustness and heterogeneity for linked interventions in the Peruvian Amazon

Abstract Large areas of forest are allocated to commercial logging concessions, some of which are eco‐certified. Theoretically, both holding concessions' rights to log and being eco‐certified can increase or decrease forest loss. Impact estimates are sparse and unsurprisingly mixed for these interventions (that aim to affect forest outcomes, if not always for conservation). We ask if adding different forest sensors, forest outcomes, and variations in forest contexts sheds new light upon the forest impacts of uncertified concessions and eco‐certifications. Using new global data, which include forest degradation, we estimate the forest impacts of these interventions within the Peruvian Amazon, disaggregating across timber regions to allow for heterogeneity in impacts across distinct contexts. Our results confirm that, when averaging across the regions, uncertified logging concessions do not increase deforestation (relative to unintervened forest), consistent with previous findings. Yet, separating regions adds insight for policy and external validity: uncertified concessions yield larger reductions in deforestation rates within the more highly pressured of these three core logging regions. Degradation data provide further insights: uncertified concessions did not shift transitions from forest to degraded status, yet reduced transitions from degraded status to deforested. Eco‐certifications did not reduce deforestation or degradation across contexts or outcomes.